This Article 
 Bibliographic References 
 Add to: 
Classification with Nonmetric Distances: Image Retrieval and Class Representation
June 2000 (vol. 22 no. 6)
pp. 583-600

Abstract—One of the key problems in appearance-based vision is understanding how to use a set of labeled images to classify new images. Classification systems that can model human performance, or that use robust image matching methods, often make use of similarity judgments that are nonmetric; but when the triangle inequality is not obeyed, most existing pattern recognition techniques are not applicable. We note that exemplar-based (or nearest-neighbor) methods can be applied naturally when using a wide class of nonmetric similarity functions. The key issue, however, is to find methods for choosing good representatives of a class that accurately characterize it. We show that existing condensing techniques for finding class representatives are ill-suited to deal with nonmetric dataspaces. We then focus on developing techniques for solving this problem, emphasizing two points: First, we show that the distance between two images is not a good measure of how well one image can represent another in nonmetric spaces. Instead, we use the vector correlation between the distances from each image to other previously seen images. Second, we show that in nonmetric spaces, boundary points are less significant for capturing the structure of a class than they are in Euclidean spaces. We suggest that atypical points may be more important in describing classes. We demonstrate the importance of these ideas to learning that generalizes from experience by improving performance using both synthetic and real images. In addition, we suggest ways of applying parametric techniques to supervised learning problems that involve a specific nonmetric distance functions, showing in particular how to generalize the idea of linear discriminant functions in a way that may be more useful in nonmetric spaces.

[1] R. Basri, L. Costa, D. Geiger, and D. Jacobs, “Determining the Similarity of Deformable Objects,” Vision Research, vol. 38, no. 15-16, pp. 2,365-2,385, 1998.
[2] M. Black and P. Anandan, The Robust Estimation of Multiple Motions: Parametric and Piecewise-Smooth Flow Fields J. Computer Vision and Image Understanding, vol. 63, no. 1, pp. 75-104, 1996.
[3] A. Blake and A. Zisserman, Visual Reconstruction. MIT Press, 1987.
[4] M. Blatt, S. Wiseman, and E. Domany, “Clustering Data through an Analogy to the Potts Model,” Advances in Neural Information Processing Systems, vol. 8, pp. 416-422, 1996.
[5] M. Brand, “A Fast Greedy Pairwise Distance Clustering Algorithm and Its Use in Discovering Thematic Structures in Large Data Sets,” Technical Report 406, MIT Media Lab, 1996.
[6] C. Cortes and V. Vapnik, "Support Vector Networks," Machine Learning, vol. 20, no. 3, Sept. 1995, pp. 1-25.
[7] I. Cox et al., "Pichunter: Bayesian Relevance Feedback for Image Retrieval," Proc. Int'l Conf. Pattern Recognition, vol. 3, Aug. 1996, pp. 362-369.
[8] B.V. Dasarathy, “Minimal Consistent Set (MCS) Identification for Optimal Nearest Neighbor Decision Systems Design,” IEEE Trans. Systems, Man, Cybernetics, vol. 24, no. 3, pp. 511-517, 1994.
[9] M. Donahue, D. Geiger, R. Hummel, and T Liu, Sparse Representations for Image Decompositions with Occlusions Proc. IEEE Conf. Computer Vision and Pattern Recognition, pp. 7-12, 1996.
[10] J.H. Friedman, J.L. Bentley, and R.A. Finkel, "An Algorithm for Finding Best Matches in Logarithmic Expected Time," ACM Trans. on Math. Software, vol. 3, no. 3, pp. 209-226, Sept. 1977.
[11] Y. Gdalyahu and D. Weinshall, “Flexible Syntactic Matching of Curves and its Application to Automatic Hierarchical Classification of Silhouettes,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 21, no. 12, pp. 1312-1328, Dec. 1999.
[12] D. Geiger and F. Girosi,“Parallel and deterministic algorithms from MRFs: Surface reconstruction,” IEEE Transactions on PAMI, vol. 13, no. 5, pp. 401-412, May 1991.
[13] S. Geman and D. Geman, “Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 6, no. 7, pp. 721-741, July 1984.
[14] K. Gowda and G. Krishna, “The Condensed Nearest Neighbor Rule Using the Concept of Mutual Nearest Neighbor,” IEEE Trans. Information Theory, vol. 25, no. 4, pp. 488-490, 1979.
[15] T. Graepel, R. Herbrich, P. Bollmann-Sdorra, and K. Obermayer, “Classification on Pairwise Proximity Data,” Proc. Neural Information Processing Systems, pp. 438-444, 1999.
[16] K. Fukunaga and J. Mantock, “Nonparametric Data Reduction,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 6, no. 1, pp. 115-118, Jan. 1984.
[17] R.M. Haralick and L.G. Shapiro, Computer and Robot Vision. New York: Addison-Wesley, 1993.
[18] P.E. Hart, “The Condensed Nearest Neighbor Rule,” IEEE Trans. Information Theory, vol. 14, no. 3, pp. 515-516, 1968.
[19] T. Hastie and W. Stuetzle, “Principal Curves,” J. Am. Statistical Assoc., vol. 84, 502-516, 1989.
[20] G. Hinton, C. Williams, and M. Revow, “Adaptive Elastic Models for Hand-Printed Character Recognition,” Neural Information Processing Systems, vol. 4, pp. 512-519, 1992.
[21] T. Hofmann and M. Buhmann, Pairwise Data Clustering by Deterministic Annealing IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 18, no. 1, pp. 1-14, Jan. 1997.
[22] D.P. Huttenlocher, G.A. Klanderman, and W.J. Rucklidge, “Comparing Images Using the Hausdorff Distance,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 15, no. 9, pp. 850-863, Sept. 1993.
[23] D.P. Huttenlocher, J.J. Noh, and W.J. Rucklidge, "Tracking Non-Rigid Objects in Complex Scenes," Proc. IEEE Int'l Conf. Computer Vision, 1993.
[24] D. Jacobs, "Linear Fitting with Missing Data," Proc. IEEE CVPR, 1997.
[25] A. Jain and D. Zongker, Feature Selection: Evaluation, Application, and Small Sample Performance IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 19, no. 2, pp. 153-158, Feb. 1997.
[26] W. Johnson and J. Lindenstrauss, “Extension of Lipshitz Mapping to Hilbert Space,” Contemporary Math., vol. 26, pp. 189-206, 1984.
[27] H. Klock and J. Buhmann, “Multidimensional Scaling by Deterministic Annealing,” Proc. Int'l Workshop Energy Minimization Methods in Computer Vision and Pattern Recognition, pp. 245-260, 1997.
[28] J. Kapur and H. Kesavan, Entropy Optimization Principles with Applications. Academic Press, 1992.
[29] S.Z. Li, "On Discontinuity-Adaptive Smoothness Priors in Computer Vision," IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 17, no. 6, pp. 576-586, June 1995.
[30] N. Linial, E. London, and Y. Rabinovich, “The Geometry of Graphs and Some of Its Algorithmic Applications,” Combinatorica, vol. 15, pp. 215-245, 1995.
[31] R. Little and D. Rubin, Statistical Analysis With Missing Data. Wiley, 1987.
[32] P. Meer, D. Mintz, and A. Rosenfeld, “Robust Regression Methods for Computer Vision: A Review,” Int'l J. Computer Vision, vol. 6, no. 1, pp. 59-70, 1991.
[33] L. Ornstein, “Computer Learning and the Scientific Method: A Proposed Solution to the Information Theoretical Problem of Meaning,” J. Mount Sinai Hospital, vol. 32, no. 4, pp. 437-494, 1965.
[34] T. Poggio and F. Girosi, “Regularization Algorithms for Learning That are Equivalent to Multilayer Networks,” Science, vol. 247, pp. 978-982, 1990.
[35] J. Puzicha, Y. Rubner, C. Tomasi, and J. Buhmann, Empirical Evaluation of Dissimilarity Measures for Color and Texture Proc. Int'l Conf. Computer Vision, 1999.
[36] H. Royden, Real Analysis. New York: MacMillan Publishing, 1968.
[37] S. Santini and R. Jain, “Similarity Queries in Image Databases,” Proc. IEEE Int'l Conf. Computer Vision and Pattern Recognition, (CVPR '96), June 1996.
[38] C. Tappert, “Cursive Script Recognition by Elastic Matching,” IBM J. Res. Development, vol. 26, no. 6 pp. 765-771, 1982.
[39] C. Tomasi and T. Kanade, "Shape and Motion From Image Streams Under Orthography: A Factorization Method," Int'l J. Computer Vision, vol. 9, no. 2, pp. 137-154, 1992.
[40] W. Tsai and S. Yu, “Attributed String Matching with Merging for Shape Recognition,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 4, pp. 453-462, July 1985.
[41] A. Tversky, “Features of Similarity,” Psychological Rev., vol. 84, no. 4, pp. 327-352, 1977.
[42] P. Williams, “Prototypes, Exemplars, and Object Recognition,” PhD thesis, Dept. of Psychology, Yale Univ., 1997.
[43] P. Yianilos, “Data Structures and Algorithms for Nearest Neighbor Search in General Metric Spaces,” Proc. Third Ann. ACM-SIAM Symp. Discrete Algorithms, pp. 311-321, 1993.
[44] K. Yoshida and H. Sakoe, “Online Handwritten Character Recognition for a Personal Computer System,” IEEE Trans. Consumer Electronics, vol. 28, no. 3, pp. 202-209, 1982.

Index Terms:
Nonmetric, image retrieval, classification, supervised learning, median, condensing, nearest-neighbor, triangle inequality, robust distance, representation.
David W. Jacobs, Daphna Weinshall, Yoram Gdalyahu, "Classification with Nonmetric Distances: Image Retrieval and Class Representation," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 22, no. 6, pp. 583-600, June 2000, doi:10.1109/34.862197
Usage of this product signifies your acceptance of the Terms of Use.